UK police reopen a Grok deepfake case as Ofcom narrows its X investigation: what it means for Tesla's in-car AI

The fallout from last winter's Grok deepfake scandal is still playing out in Britain. South Wales Police have reopened the case of Jess Davies, a woman whose sexually explicit AI-generated image was posted on X, the Bureau of Investigative Journalism reported on October 22. A week earlier, on October 14, the media regulator Ofcom published an update on its investigation into X.

What happened in the Davies case

Around Christmas 2025, explicit images made with Grok, the AI built by xAI (now SpaceXAI), were being posted on X at a rate of up to 6,000 an hour, according to the Bureau. Davies reported an image of herself to the police. Her case was closed in May 2026 because officers could not identify the user behind it.

Police had treated the image under malicious communications law rather than under sexual offences legislation, which covers deepfakes. After the Bureau's reporting, journalists passed new information about the user to the police, including a crypto wallet and dozens of deleted posts. Police are now investigating under sexual offences law and making enquiries with X. Davies said she was pleased, but called it frustrating that victims have to fight to be taken seriously. She is also starting civil legal action against X.

SpaceXAI did not respond to the Bureau's request for comment.

What Ofcom says has changed

Ofcom opened its investigation into X in January 2026, after reports that the Grok account on X was being used to create and share sexual deepfakes of real people, including children. In its October update, it said:

• X: now consistently restricts content labelled as pornography for accounts that have not passed age checks. Because of that, Ofcom has narrowed the rest of its investigation to X's duties on illegal content and on complaints and reporting.
• Standalone Grok: Ofcom said gaps in the Online Safety Act's coverage of chatbots meant it could not investigate illegal image creation in the separate Grok app. After talks with the regulator, SpaceXAI blocked the generation of images of real people in revealing clothing, stopped UK users from viewing, generating or sharing pornographic content through the image tool, and stopped companion chatbots such as "Ani" from sexually explicit conversations with UK users.
• Next steps: Ofcom will decide whether X complied with its duties. If it believes X failed, it will issue a provisional view, and X can respond before a final decision.

The Bureau, however, found that users could still request sexualised images of women, such as bikini edits, without consent, including through the standalone Grok app.

Why this matters for Tesla

Grok is the voice assistant in Tesla cars, and Tesla's AI ambitions, from the in-car assistant to conversation with the Optimus robot, lean on the same company. That makes Grok's reputation part of Tesla's brand risk, especially in Europe and the UK, where regulators are watching AI closely.

It is important to keep the two apart, though. According to Tesla's support page, Grok in the car is built around driving and daily tasks: navigation, calls, music, climate, questions about the car, and with extra features email, calendar and multi-step tasks such as reservations. Image generation is not among the in-car features Tesla lists. Tesla decides through its own software what the assistant can do in the car.

Tesla also has a financial tie: its $2 billion investment in xAI was converted into a small stake in SpaceX. Legal and regulatory costs at SpaceXAI therefore touch Tesla shareholders, if only slightly.

What to watch next

• Whether Ofcom issues a provisional finding against X, and any penalties that follow.
• Changes to UK law to cover standalone AI chatbots, which Ofcom says it cannot fully police today.
• The outcome of the reopened police case and Davies's civil claim.